CN117828483A - A fault prediction analysis method and system for building equipment data fusion - Google Patents

Abstract

The present application provides a method and system for fault prediction and analysis of building equipment data fusion, which relates to the field of data processing technology. The method comprises: collecting first real-time operating data of a first building equipment and second real-time operating data of a second building equipment; obtaining collaborative fault characteristics; obtaining a first independent fault probability and a second independent fault probability; performing probability fusion calculations and outputting a fused fault probability; when the fused fault probability is greater than a preset fused fault probability, outputting fault warning information. This solves the technical problem that most of the fault prediction and analysis in the prior art is performed separately for a single device, and there is a lack of collaborative fault analysis between collaborative devices, which leads to incomplete and inaccurate fault warnings. By performing fault probability analysis and prediction on collaborative equipment and performing fault warnings in a timely manner, the technical effect of ensuring the normal execution of collaborative operations and improving the accuracy and sensitivity of fault warnings is achieved.

Description

A fault prediction analysis method and system for building equipment data fusion

Technical Field

The present application relates to the field of data processing technology, and in particular to a method and system for fault prediction analysis of building equipment data fusion.

Background Art

Building equipment engineering is an important part of modern buildings. It covers all aspects of a building, such as water supply, drainage, ventilation, air conditioning, fire protection, electrical lighting, etc. It is directly related to the safety, comfort, and energy consumption of the building. This places a high demand on the safety of building equipment. After the building equipment is put into operation, it is prone to failure as the operating time increases. Among the building equipment currently used, two or more devices are often required to cooperate with each other to complete collaborative tasks, such as environmental sensors and controllers. If the equipment operates abnormally during the cooperation process, the collaborative task will be interrupted and cannot be executed. However, in the existing technology, most of them are fault prediction and analysis of a single device separately, lacking collaborative fault analysis between collaborative devices, which leads to incomplete and inaccurate fault warning.

Summary of the invention

The present application provides a fault prediction analysis method and system for building equipment data fusion, which is used to solve the technical problem that most of the fault prediction analysis in the prior art is performed separately on a single device, lacking collaborative fault analysis between collaborative devices, resulting in incomplete and inaccurate fault warnings.

According to the first aspect of the present application, a fault prediction and analysis method for building equipment data fusion is provided, including: collecting first real-time operating condition data of a first building equipment and second real-time operating condition data of a second building equipment, wherein the first building equipment and the second building equipment are collaborative working devices; performing collaborative working abnormality analysis based on the first real-time operating condition data and the second real-time operating condition data to obtain collaborative fault characteristics; inputting the collaborative fault characteristics into a fault prediction module to perform fault prediction on the first real-time operating condition data and the second real-time operating condition data respectively, to obtain a first independent fault probability indicating that the first building equipment has failed, and a second independent fault probability indicating that the second building equipment has failed; performing probability fusion calculation based on the first independent fault probability and the second independent fault probability, and outputting a fused fault probability; when the fused fault probability is greater than a preset fused fault probability, outputting a fault warning message.

According to a second aspect of the present application, a fault prediction and analysis system for building equipment data fusion is provided, including: a real-time working condition data acquisition unit, the real-time working condition data acquisition unit is used to collect first real-time working condition data of a first building equipment and second real-time working condition data of a second building equipment, wherein the first building equipment and the second building equipment are collaborative working devices; a collaborative working abnormality analysis unit, the collaborative working abnormality analysis unit is used to perform collaborative working abnormality analysis based on the first real-time working condition data and the second real-time working condition data, and obtain collaborative fault characteristics; an independent fault probability prediction unit, the independent fault probability prediction unit is used to input the collaborative fault characteristics into a fault prediction module to perform fault prediction on the first real-time working condition data and the second real-time working condition data respectively, and obtain a first independent fault probability indicating that the first building equipment has failed, and a second independent fault probability indicating that the second building equipment has failed; a probability fusion calculation unit, the probability fusion calculation unit is used to perform probability fusion calculation based on the first independent fault probability and the second independent fault probability, and output a fused fault probability; a fault reminder unit, the fault reminder unit is used to output fault reminder information when the fused fault probability is greater than a preset fused fault probability.

According to one or more technical solutions adopted in this application, the beneficial effects that can be achieved are as follows:

The first real-time working condition data of the first construction equipment and the second real-time working condition data of the second construction equipment are collected, wherein the first construction equipment and the second construction equipment are collaborative working equipment, and the collaborative working abnormality analysis is performed according to the first real-time working condition data and the second real-time working condition data to obtain collaborative fault characteristics, and the collaborative fault characteristics are input into the fault prediction module to perform fault prediction on the first real-time working condition data and the second real-time working condition data respectively, and the first independent fault probability indicating that the first construction equipment has failed, and the second independent fault probability indicating that the second construction equipment has failed are obtained, and the probability fusion calculation is performed according to the first independent fault probability and the second independent fault probability, and the fusion fault probability is output, and when the fusion fault probability is greater than the preset fusion fault probability, the fault reminder information is output. In this way, by analyzing and predicting the fault probability of the collaborative working equipment and timely carrying out fault warning, the technical effect of ensuring the normal execution of the collaborative working and improving the accuracy and sensitivity of the fault warning is achieved.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in this application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. The drawings constituting a part of this application are used to provide a further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation of this application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without creative work.

FIG1 is a schematic diagram of a flow chart of a method for fault prediction and analysis of building equipment data fusion provided in an embodiment of the present application;

FIG2 is a schematic diagram of the structure of a fault prediction and analysis system for building equipment data fusion provided in an embodiment of the present application.

Explanation of the accompanying drawings: real-time working condition data acquisition unit 11, collaborative operation abnormality analysis unit 12, independent fault probability prediction unit 13, probability fusion calculation unit 14, fault reminder unit 15.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present application more obvious, the exemplary embodiments of the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

The terms used in the specification are used to describe the embodiments, rather than to limit the present application. As used in the specification, the singular terms "a", "an" and "the" are intended to also include plural forms, unless the context clearly indicates otherwise. When used in the specification, the terms "comprise" and/or "include" specify the presence of steps, operations, elements and/or components, but do not exclude the presence or addition of one or more other steps, operations, elements, components and/or groups thereof.

Unless otherwise defined, all terms (including technical and scientific terms) used in this specification shall have the same meaning as commonly understood by those skilled in the art to which this application belongs. Terms, such as those defined in commonly used dictionaries, should not be interpreted in an idealized or overly formal sense unless explicitly defined herein. Throughout the specification, the same reference numerals represent the same elements.

It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

Embodiment 1:

FIG1 is a diagram of a method for fault prediction and analysis of building equipment data fusion provided by an embodiment of the present application, the method comprising:

Collecting first real-time working condition data of a first construction equipment and second real-time working condition data of a second construction equipment, wherein the first construction equipment and the second construction equipment are collaboratively operated equipment;

The first building equipment and the second building equipment are equipment, modules or systems that need to cooperate and collaborate with each other to complete a task in any scenario, such as environmental sensors and controllers. Then, the first real-time working condition data of the first building equipment and the second real-time working condition data of the second building equipment are collected. The first real-time working condition data and the second real-time working condition data are the real-time operating parameters of the first building equipment and the second building equipment, such as the energy consumption, heat, load, etc. of the environmental sensor. The data types of the first real-time working condition data and the second real-time working condition data can be determined by professional and technical personnel in this field, and then sensors in the prior art are installed on the first building equipment and the second building equipment, and the first real-time working condition data and the second real-time working condition data are obtained through the sensors.

Performing collaborative operation abnormality analysis based on the first real-time operating condition data and the second real-time operating condition data to obtain collaborative fault characteristics;

According to the first real-time working condition data and the second real-time working condition data, the collaborative operation abnormality analysis is performed to obtain the collaborative fault characteristics. The collaborative fault characteristics refer to the data characteristics of abnormal deviation when the first building equipment and the second building equipment work together. For example, one of the first real-time working condition data and the second real-time working condition data has a large deviation, resulting in the failure of the collaborative task between the first building equipment and the second building equipment to be completed normally. At this time, the working condition data is the collaborative fault characteristics. Exemplarily, in the collaborative work of the environmental sensor and the controller, the environmental sensor is responsible for collecting various data about the surrounding environment, such as temperature, humidity, light, air quality, etc., and the controller is responsible for receiving the data from the sensor and processing it according to preset rules or algorithms, such as adjusting the operating state of the system or issuing control instructions according to changes in the environmental state to achieve control of the environment. If the environmental sensor operates abnormally, such as abnormal power consumption, which will cause a large error in the sensor data or fail to collect environmental data, it will make it difficult for the controller to accurately control the environmental conditions; similarly, if the controller is abnormal, it will also cause the entire environmental control process to be unable to be accurately implemented. Simply put, based on the existing technology, the normal operating condition data corresponding to the normal operation of the first construction equipment and the second construction equipment can be obtained, and then the deviation between the first real-time operating condition data and the second real-time operating condition data and the normal data can be compared, and the data with abnormal deviation can be used as a collaborative fault feature.

In a preferred embodiment, it also includes:

Obtain historical collaborative working condition data of the first construction equipment and the second construction equipment, perform collaborative operation anomaly analysis on the first real-time working condition data and the second real-time working condition data based on the historical collaborative working condition data, and obtain collaborative fault characteristics; wherein the collaborative fault characteristics are composed of indicators with a data deviation greater than or equal to a preset deviation.

Specifically, the historical collaborative working condition data of the first construction equipment and the second construction equipment are obtained. The historical collaborative working condition data refers to the working condition parameters of the first construction equipment and the second construction equipment when they cooperated in the past period of time, such as equipment working condition load, equipment working condition energy consumption and equipment working condition heat, etc. It should be noted that the historical collaborative working condition data is obtained by sampling the historical working conditions of the first construction equipment and the second construction equipment. Based on the historical collaborative working condition data, the first real-time working condition data and the second real-time working condition data are analyzed for collaborative operation anomalies, that is, the historical collaborative working condition data and the first real-time working condition data and the second real-time working condition data are analyzed for data deviation respectively, and the data with abnormal deviation is used as a collaborative fault feature. When the number of collaborative fault features is large, the collaborative fault features can be further screened, and the indicators with data deviation greater than or equal to the preset deviation constitute the collaborative fault features. The preset deviation is set by professional and technical personnel in this field, and the deviation allowed in the actual operation process needs to be set in combination with actual experience. In layman's terms, the deviations between the historical collaborative working condition data and the first real-time working condition data and the second real-time working condition data are compared, and the degree of deviation between the first real-time working condition data and the data in the historical collaborative working condition data is calculated. If the degree of deviation is greater than or equal to the predetermined deviation, the corresponding data is used as a collaborative fault feature. This enables the analysis of collaborative fault features, facilitates subsequent collaborative fault warnings, and prevents collaborative equipment from malfunctioning and causing construction operation errors.

Input the collaborative fault feature into a fault prediction module to perform fault prediction on the first real-time operating condition data and the second real-time operating condition data, respectively, to obtain a first independent fault probability indicating a fault of the first building equipment and a second independent fault probability indicating a fault of the second building equipment;

The fault prediction module is constructed based on the machine learning model in the prior art. That is to say, after the collaborative fault features are obtained, the collaborative fault features can be input into the fault prediction module to perform fault prediction on the first real-time working condition data and the second real-time working condition data. The fault prediction module can use existing machine learning, statistical analysis and other algorithms and techniques to process and analyze the input collaborative fault features and the first real-time working condition data and the second real-time working condition data to predict the probability of failure of the first building equipment and the second building equipment.

In a preferred embodiment, it also includes:

Collect historical fault characteristics of the first construction equipment; collect historical fault characteristics of the second construction equipment; if the historical fault characteristics of the first construction equipment include the collaborative fault characteristics, and the historical fault characteristics of the second construction equipment include the collaborative fault characteristics; collect the first interval duration of the first construction equipment, collect the second interval duration of the second construction equipment, wherein the interval duration is the interval duration from the last failure of the equipment; according to the fault prediction module, predict the abnormal operating condition probability of the first construction equipment based on the first interval duration and the collaborative fault characteristics, and output it as a first independent fault probability, and predict the abnormal operating condition probability of the second construction equipment based on the second interval duration and the collaborative fault characteristics, and output it as a second independent fault probability.

Specifically, the historical fault characteristics of the first building equipment are collected, that is, the fault records of the first building equipment in the past period of time are retrieved, and the operating parameters and the corresponding fault time when the fault occurs are extracted as the historical fault characteristics. Similarly, the operating parameters and the corresponding fault time when the second building equipment fails in the past period of time are collected as the historical fault characteristics of the second building equipment. The historical fault characteristics of the first building equipment and the historical fault characteristics of the second building equipment are compared with the collaborative fault characteristics respectively. If the historical fault characteristics of the first building equipment include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment include the collaborative fault characteristics, the first interval duration of the first building equipment is collected, and the second interval duration of the second building equipment is collected, wherein the first interval duration refers to the time length from the last occurrence of the collaborative fault of the first building equipment to the present, and the time length from the last occurrence of the collaborative fault characteristics to the present can be obtained based on the historical fault characteristics of the first building equipment as the first interval duration; similarly, the second interval duration refers to the time length from the last occurrence of the collaborative fault of the second building equipment to the present.

Specifically, based on the statistical analysis of the historical fault characteristics of the first building equipment, the occurrence time of the corresponding collaborative fault characteristics is obtained to obtain the fault occurrence duration, that is, how often the collaborative fault occurs, and then compare the deviation between the fault occurrence duration and the first interval duration. The smaller the deviation, the greater the first independent fault probability. Exemplarily, the ratio of the first interval duration to the fault occurrence duration can be used as the first independent fault probability. Similarly, the same method as that for obtaining the first independent fault probability is used to predict the second independent fault probability of the second building equipment under the second interval duration. The above is the process of fault prediction through the fault prediction module. In this way, the independent fault probability prediction of the first building equipment and the second building equipment is achieved, providing support for the subsequent fault probability fusion.

Perform probability fusion calculation according to the first independent fault probability and the second independent fault probability, and output a fused fault probability;

In a preferred embodiment, it also includes:

A probability fusion calculation is performed according to the first independent fault probability and the second independent fault probability, and a fusion fault probability is output. The expression of the probability fusion calculation includes:

,

in, First building equipment and second building equipment The probability of fusion failure, for the first construction equipment based on the first interval duration and the first independent failure probability of the collaborative failure signature D analysis, Where D is the collaborative fault characteristic A set of n is the number of features in the set, is the weight corresponding to the first independent fault probability obtained through integrated fusion model training;

is a second independent failure probability of the second building equipment based on the second interval duration and the coordinated failure feature, is the weight corresponding to the second independent fault probability obtained through integrated fusion model training.

In a preferred embodiment, it also includes:

If the historical failure characteristics of the first building equipment include the collaborative failure characteristics, and the historical failure characteristics of the second building equipment do not include the collaborative failure characteristics; predict the first independent failure probability of the first building equipment under the first interval time condition; obtain the second initial failure probability corresponding to the second building equipment, and perform probability fusion calculation with the first independent failure probability and the second initial failure probability. The expression of probability fusion calculation includes:

,

in, is a second initial failure probability corresponding to the second building equipment based on the collaborative failure feature, is the weight corresponding to the second initial fault probability obtained through integrated fusion model training.

In a preferred embodiment, it also includes:

If the historical fault characteristics of the first building equipment do not include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment include the collaborative fault characteristics; predict the second independent fault probability of the second building equipment under the second interval time condition; obtain the first initial fault probability corresponding to the first building equipment, and perform probability fusion calculation with the first initial fault probability and the second independent fault probability. The expression of probability fusion calculation includes:

,

in, is a first initial failure probability corresponding to the first building equipment based on the collaborative failure feature, is the weight corresponding to the first initial fault probability obtained through integrated fusion model training.

Specifically, a probability fusion calculation is performed according to the first independent fault probability and the second independent fault probability, and a fusion fault probability is output. In the expression of the probability fusion calculation, Based on the first building equipment and second building equipment The probability of fusion failure, for the first construction equipment based on the first interval duration and the first independent failure probability of the collaborative failure signature D analysis, Where D is the collaborative fault characteristic A set of features, n is the number of features in the set. Exemplarily, the collaborative fault feature D may include operating condition features such as equipment operating load, equipment operating energy consumption and equipment operating heat. In the aforementioned steps, the occurrence time of the corresponding collaborative fault feature is statistically analyzed based on the historical fault features of the first building equipment to obtain the fault occurrence duration, and the ratio of the first interval duration to the fault occurrence duration is used as the first independent fault probability. The collaborative fault feature may include multiple different types of fault features to achieve a comprehensive analysis of collaborative faults. is the weight corresponding to the first independent fault probability obtained through integrated fusion model training; is a second independent failure probability of the second building equipment based on the second interval duration and the coordinated failure feature, is the weight corresponding to the second independent failure probability obtained through integrated fusion model training. Simply put, the fused failure probability is a way of weighted fusion of the first independent failure probability and the second independent failure probability as two probability features, that is, the weights corresponding to different collaborative devices are obtained through integrated fusion model training according to the degree of influence of the independent failure probability of different collaborative devices on the fused failure probability. For example, when an environmental sensor fails, the impact on the collaborative operation between the environmental sensor and the controller is greater, then its impact on the fused failure probability is also greater, and the corresponding weight obtained from training is also greater, such as 0.8. Specifically, the weight and The training acquisition can be performed by collecting the failure probability training sample set of the first construction equipment, the failure probability training sample set of the second construction equipment, and the failure probability verification sample set of the collaborative operation of the first construction equipment and the second construction equipment. Then, the first independent failure probability and the second independent failure probability are weightedly calculated, and the weighted calculation result is the fused failure probability, thereby realizing the fusion of failure probability and improving the accuracy of fault warning for collaborative equipment. It should be noted that the above expression is only used when the historical failure characteristics of the first construction equipment include the collaborative failure characteristics, and the historical failure characteristics of the second construction equipment include the collaborative failure characteristics. Information on the failure of the same task target is provided through two or more sources of input data, and can be integrated to enhance the credibility of the probabilistic warning. The prior art mostly uses the analysis of early warning for a single device. When two or more collaborative devices are performing a certain task, if a separate early warning analysis is performed on each device separately, the risk of the task will be falsely reported or missed, which makes the task have operational risks. Therefore, the scheme proposed in this application is for two or more collaborative devices to perform the same task. In order to analyze the risk of the task, it is necessary to consider the risk of failure of each device at present, and integrate the failure risk of each device to output the risk of the corresponding task, and then make an alarm judgment on the fused indicator according to the preset fusion failure probability, so as to reduce the missed reports and false alarms caused by the failure of a single device in the collaborative operation scenario.

For example, if the first building equipment is an environmental sensor and the second building equipment is a controller, when the probability of a single failure of any building equipment is greater than the warning threshold of 0.26, a warning signal can be issued separately for equipment management. When both buildings are not in the warning range, the method of the present application is implemented. Specifically, if the probability of the environmental sensor is obtained respectively , the probability of the controller In the prior art, the probability of the device , The values of the environmental sensors and controllers are all less than the threshold of 0.26, so it is determined that no warning signal is issued, and a missed alarm occurs. The solution of this application first optimizes the weights of the environmental sensors and controllers to obtain , assuming , then and Perform probability fusion calculation and substitute the above expression to get the calculation formula: , and then make a warning judgment based on 0.222, and judge it with the preset fusion failure probability (0.2) configured in advance. At this time, 0.222＞0.2, and then issue a warning signal.

If the historical fault characteristics of the first building equipment include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment do not include the collaborative fault characteristics, the aforementioned method is used to predict the first independent fault probability of the first building equipment under the first interval time condition. Further obtain the second initial fault probability corresponding to the second building equipment, that is, when the historical fault characteristics of the first building equipment include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment do not include the collaborative fault characteristics, it is still necessary to perform collaborative fault probability fusion. At this time, calculate the probability of the first occurrence of the collaborative fault characteristics in the second building equipment. In layman's terms, the second building equipment has not had the type of failure corresponding to the collaborative fault characteristics in history, so the probability of the first occurrence based on the collaborative fault characteristics is calculated. Specifically, based on the existing technology, the same family of building equipment with the same equipment type and operating time as the second building equipment can be retrieved, and the probability of the first occurrence of the type of failure corresponding to the collaborative fault characteristics of the same family of building equipment based on the existing technology is used as the second initial failure probability.

Then, a probability fusion calculation is performed using the first independent fault probability and the second initial fault probability. The expression of the probability fusion calculation includes:

,

in, is the second initial failure probability of the second building equipment corresponding to the collaborative failure feature, that is, the second building equipment has not experienced the failure type corresponding to the collaborative failure feature in history, so the probability of the first occurrence of the failure type corresponding to the collaborative failure feature is calculated; is the weight corresponding to the second initial fault probability obtained by training the integrated fusion model.

If the historical fault characteristics of the first building equipment do not include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment include the collaborative fault characteristics, the second independent fault probability of the second building equipment under the second interval time condition is predicted by the aforementioned fault prediction module. The first initial fault probability corresponding to the first building equipment is further obtained. The first initial fault probability is obtained in the same way as the second initial fault probability, that is, the first building equipment has not had the type of fault corresponding to the collaborative fault characteristics in history, so the probability of the first occurrence of the type of fault corresponding to the collaborative fault characteristics is counted. Specifically, based on the existing technology, the same family of building equipment as the first building equipment can be retrieved, and the probability of the first occurrence of the type of fault corresponding to the collaborative fault characteristics of the same family of building equipment is counted as the first initial fault probability.

The probability fusion calculation is performed using the first initial fault probability and the second independent fault probability. The expression of the probability fusion calculation includes:

,

in, The first initial failure probability corresponding to the collaborative failure feature of the first building equipment is that the first building equipment has not had a failure type corresponding to the collaborative failure feature in history, so the probability of the first occurrence of the failure type corresponding to the collaborative failure feature is calculated. The weight corresponding to the first initial failure probability obtained through integrated fusion model training is used. When the failure type corresponding to the collaborative failure feature does not occur in any of the first building equipment and the second building equipment, the probability of the failure type corresponding to the collaborative failure feature occurring for the first time is statistically calculated, and then the failure probability is fused and calculated to improve the comprehensiveness of the collaborative failure probability prediction.

In a preferred embodiment, it also includes:

Obtain m initial candidate fusion models, and assign m different weight ratios to the m initial candidate fusion models ; Construct a fusion sample set, the fusion sample includes a failure probability training sample set based on the first construction equipment, a failure probability training sample set of the second construction equipment, and a failure probability verification sample set of the collaborative operation of the first construction equipment and the second construction equipment; train the m initial candidate fusion models according to the fusion sample set, and output an integrated fusion model.

In the above expression of probability fusion calculation, is the weight corresponding to the first independent fault probability obtained through integrated fusion model training, is the weight corresponding to the second independent fault probability obtained through integrated fusion model training, where and The acquisition process is as follows:

Obtain m initial candidate fusion models, and assign m different weight ratios to the m initial candidate fusion models , m is an integer greater than 1, and the initial candidate fusion model is a machine learning model in the prior art, which is used to weight the independent failure probabilities of the first construction equipment and the second construction equipment according to the weight ratio. Construct a fusion sample set, and the fusion sample includes a failure probability training sample set based on the first construction equipment, a failure probability training sample set of the second construction equipment, and a failure probability verification sample set of the collaborative operation of the first construction equipment and the second construction equipment. In layman's terms, the data in the fusion sample set can be understood as historical data, and the sample data in the failure probability training sample set of the first construction equipment, the failure probability training sample set of the second construction equipment, and the failure probability verification sample set of the collaborative operation of the first construction equipment and the second construction equipment have a one-to-one correspondence. Specifically, it can be obtained by retrieving the failure records of the first construction equipment and the second construction equipment in the history in the prior art, or by retrieving the failure records of the same collaborative operation equipment as the first construction equipment and the second construction equipment. Then, the m initial candidate fusion models are trained according to the fusion sample set, that is, the data with corresponding relationship in the failure probability training sample set of the first building equipment and the failure probability training sample set of the second building equipment are input into the m initial candidate fusion models, and the m initial candidate fusion models use their corresponding weight ratios to perform weighted calculations and output fusion probabilities. The output fusion probabilities are analyzed for accuracy according to the failure probability verification sample set, and the weight ratios of the m initial candidate fusion models are adjusted until the accuracy meets the requirements, and then the m initial candidate fusion models whose accuracy meets the requirements are averagely weighted to obtain an integrated fusion model. In this way, the construction of the integrated fusion model is realized, and the integrated fusion model is obtained. and , improving the accuracy of fault fusion analysis.

In a preferred embodiment, it also includes:

The m initial candidate fusion models are trained according to the fault probability training sample set of the first building equipment and the fault probability training sample set of the second building equipment to obtain m fault probability output sample sets; based on the m fault probability output sample sets, compared with the fault probability verification sample set, an optimized candidate fusion model is obtained, wherein the probability accuracy of the optimized candidate fusion model is greater than or equal to the preset accuracy; the weight ratio of the next round is adjusted according to the optimized candidate fusion model, and so on, to obtain m candidate fusion models after the preset iteration rounds; the m candidate fusion models are weighted summed to obtain an integrated fusion model, and then the weight ratio of the integrated fusion model is output as and .

Specifically, the m initial candidate fusion models are trained according to the fault probability training sample set of the first building equipment and the fault probability training sample set of the second building equipment, that is, the data in the fault probability training sample set of the first building equipment and the fault probability training sample set of the second building equipment are input into the m initial candidate fusion models, and m fault probability output sample sets are output. The fault probability verification sample set is the output expectation of the m initial candidate fusion models, and the accuracy rates corresponding to the m fault probability output sample sets are obtained based on the comparison between the m fault probability output sample sets and the fault probability verification sample set, and the initial candidate fusion model with an accuracy rate greater than or equal to the preset accuracy rate is obtained as the optimized candidate fusion model. The preset accuracy rate is set by professional and technical personnel in this field, such as 90%, thereby obtaining the optimized candidate fusion model.

The weight ratio of the next round is adjusted according to the optimized candidate fusion model, that is, the weight ratio corresponding to the optimized candidate fusion model is used as the adjustment direction to adjust the weight ratio of the m candidate fusion models, and then the adjusted m candidate fusion models are trained again according to the failure probability training sample set of the first building equipment and the failure probability training sample set of the second building equipment, and the optimized candidate fusion model is continued to be obtained and the weight ratio of the m candidate fusion models is adjusted, and so on, to obtain m candidate fusion models after the preset iteration rounds, wherein the preset iteration rounds are set by professional and technical personnel in this field, such as 50 times, or the iteration can be ended in other ways, for example, when the accuracy of the adjusted m candidate fusion models reaches the preset accuracy, the iteration can also be stopped. Finally, the weighted sum of the m candidate fusion models finally obtained is performed to obtain an integrated fusion model, and then the weight ratio of the integrated fusion model is output as and , providing support for probability fusion calculation, thereby improving the accuracy of fusion probability analysis and fault warning accuracy.

When the fusion failure probability is greater than the preset fusion failure probability, a failure reminder message is output.

The preset fusion failure probability is set by the professional and technical personnel in this field based on actual experience, and there is no restriction on this. When the fusion failure probability is greater than the preset fusion failure probability, a fault reminder message is output to remind the operator that the first construction equipment and the second construction equipment may have a collaborative operation failure, assisting the operator to perform fault inspection and troubleshooting in a timely manner to reduce the safety risk of construction.

Based on the above analysis, it can be seen that the one or more technical solutions provided in this application can achieve the following beneficial effects:

The first real-time working condition data of the first construction equipment and the second real-time working condition data of the second construction equipment are collected, wherein the first construction equipment and the second construction equipment are collaborative working equipment, and the collaborative working abnormality analysis is performed according to the first real-time working condition data and the second real-time working condition data to obtain collaborative fault characteristics, and the collaborative fault characteristics are input into the fault prediction module to perform fault prediction on the first real-time working condition data and the second real-time working condition data respectively, and the first independent fault probability indicating that the first construction equipment has failed, and the second independent fault probability indicating that the second construction equipment has failed are obtained, and the probability fusion calculation is performed according to the first independent fault probability and the second independent fault probability, and the fusion fault probability is output, and when the fusion fault probability is greater than the preset fusion fault probability, the fault reminder information is output. In this way, by analyzing and predicting the fault probability of the collaborative working equipment and timely carrying out fault warning, the technical effect of ensuring the normal execution of the collaborative working and improving the accuracy and sensitivity of the fault warning is achieved.

Embodiment 2:

Based on the same inventive concept as the method for fault prediction and analysis of building equipment data fusion in the aforementioned embodiment, as shown in FIG2 , the present application also provides a fault prediction and analysis system of building equipment data fusion, the system comprising:

A real-time working condition data acquisition unit 11, wherein the real-time working condition data acquisition unit 11 is used to collect first real-time working condition data of a first construction device and second real-time working condition data of a second construction device, wherein the first construction device and the second construction device are collaborative working devices;

A collaborative operation abnormality analysis unit 12, the collaborative operation abnormality analysis unit 12 is used to perform collaborative operation abnormality analysis according to the first real-time working condition data and the second real-time working condition data to obtain collaborative fault characteristics;

An independent fault probability prediction unit 13, the independent fault probability prediction unit 13 is used to input the collaborative fault feature into a fault prediction module to perform fault prediction on the first real-time operating condition data and the second real-time operating condition data respectively, and obtain a first independent fault probability indicating that the first building equipment has failed, and a second independent fault probability indicating that the second building equipment has failed;

A probability fusion calculation unit 14, wherein the probability fusion calculation unit 14 is used to perform probability fusion calculation according to the first independent fault probability and the second independent fault probability, and output a fusion fault probability;

The fault reminder unit 15 is used to output fault reminder information when the fusion fault probability is greater than a preset fusion fault probability.

Furthermore, the independent fault probability prediction unit 13 also includes:

collecting historical fault characteristics of the first construction equipment;

collecting historical fault characteristics of the second building equipment;

If the historical fault characteristics of the first construction equipment include the collaborative fault characteristics, and the historical fault characteristics of the second construction equipment include the collaborative fault characteristics;

Collecting a first interval duration of the first construction equipment and collecting a second interval duration of the second construction equipment, wherein the interval duration is the interval duration from the last failure of the equipment;

According to the fault prediction module, the probability of abnormal operating condition of the first building equipment is predicted based on the first interval duration and the collaborative fault characteristics, and the output is a first independent fault probability. The probability of abnormal operating condition of the second building equipment is predicted based on the second interval duration and the collaborative fault characteristics, and the output is a second independent fault probability.

Furthermore, the probability fusion calculation unit 14 also includes:

A probability fusion calculation is performed according to the first independent fault probability and the second independent fault probability, and a fusion fault probability is output. The expression of the probability fusion calculation includes:

,

in, First building equipment and second building equipment The probability of fusion failure, for the first construction equipment based on the first interval duration and the first independent failure probability of the collaborative failure signature D analysis, Where D is the collaborative fault feature A set of n is the number of features in the set, is the weight corresponding to the first independent fault probability obtained through integrated fusion model training;

is a second independent failure probability of the second building equipment based on the second interval duration and the coordinated failure feature, is the weight corresponding to the second independent fault probability obtained through integrated fusion model training.

Furthermore, the probability fusion calculation unit 14 also includes:

If the historical fault characteristics of the first building equipment include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment do not include the collaborative fault characteristics;

Predicting a first independent failure probability of the first construction equipment under the first interval time condition;

A second initial failure probability corresponding to the second building equipment is obtained, and a probability fusion calculation is performed using the first independent failure probability and the second initial failure probability. The expression of the probability fusion calculation includes:

,

in, is a second initial failure probability corresponding to the second building equipment based on the collaborative failure feature, is the weight corresponding to the second initial fault probability obtained through integrated fusion model training.

Furthermore, the probability fusion calculation unit 14 also includes:

If the historical fault characteristics of the first building equipment do not include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment include the collaborative fault characteristics;

predicting a second independent failure probability of the second building equipment under the second interval time condition;

A first initial failure probability corresponding to the first building equipment is obtained, and a probability fusion calculation is performed using the first initial failure probability and the second independent failure probability. The expression of the probability fusion calculation includes:

,

in, is a first initial failure probability corresponding to the first building equipment based on the collaborative failure feature, is the weight corresponding to the first initial fault probability obtained through integrated fusion model training. Further, the collaborative operation abnormality analysis unit 12 also includes:

Acquire historical collaborative working condition data of the first construction equipment and the second construction equipment, and perform collaborative operation abnormality analysis on the first real-time working condition data and the second real-time working condition data based on the historical collaborative working condition data to acquire collaborative fault characteristics;

The collaborative fault feature is composed of an indicator whose data deviation is greater than or equal to a preset deviation.

Furthermore, the probability fusion calculation unit 14 also includes:

Obtain m initial candidate fusion models, and assign m different weight ratios to the m initial candidate fusion models ;

Constructing a fusion sample set, the fusion sample comprising a failure probability training sample set based on the first construction equipment, a failure probability training sample set based on the second construction equipment, and a failure probability verification sample set of the first construction equipment and the second construction equipment working in collaboration;

The m initial candidate fusion models are trained according to the fusion sample set, and an integrated fusion model is output.

Furthermore, the probability fusion calculation unit 14 also includes:

Training the m initial candidate fusion models according to the failure probability training sample set of the first building equipment and the failure probability training sample set of the second building equipment to obtain m failure probability output sample sets;

Based on the comparison of the m fault probability output sample sets with the fault probability verification sample set, an optimized candidate fusion model is obtained, wherein the probability accuracy of the optimized candidate fusion model is greater than or equal to a preset accuracy;

The weight ratio of the next round is adjusted according to the optimized candidate fusion model, and so on, to obtain m candidate fusion models after a preset number of iteration rounds;

The m candidate fusion models are weighted and summed to obtain an integrated fusion model, and then the weight ratio of the integrated fusion model is output as and .

The specific example of a fault prediction analysis method for building equipment data fusion in the aforementioned embodiment 1 is also applicable to a fault prediction analysis system for building equipment data fusion in this embodiment. Through the aforementioned detailed description of a fault prediction analysis method for building equipment data fusion, technical personnel in this field can clearly understand a fault prediction analysis system for building equipment data fusion in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here.

It should be understood that various forms of the processes shown above can be used to reorder, add or delete steps, as long as the expected results of the technical solutions disclosed in this application can be achieved, and this document does not limit them here.

Note that the above are only preferred embodiments of the present application and the technical principles used. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present application. Therefore, although the present application is described in more detail through the above embodiments, the present application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the appended claims.

Claims (9)

1. A method for fault prediction and analysis based on building equipment data fusion, characterized in that the method comprises: Collecting first real-time working condition data of a first construction equipment and second real-time working condition data of a second construction equipment, wherein the first construction equipment and the second construction equipment are collaboratively operated equipment; Performing collaborative operation abnormality analysis based on the first real-time operating condition data and the second real-time operating condition data to obtain collaborative fault characteristics; Input the collaborative fault feature into a fault prediction module to perform fault prediction on the first real-time operating condition data and the second real-time operating condition data, respectively, to obtain a first independent fault probability indicating a fault of the first building equipment and a second independent fault probability indicating a fault of the second building equipment; Perform probability fusion calculation according to the first independent fault probability and the second independent fault probability, and output a fused fault probability; When the fusion failure probability is greater than the preset fusion failure probability, a failure reminder message is output. 2. The method according to claim 1, characterized in that the collaborative fault feature is input into a fault prediction module to perform fault prediction on the first real-time operating condition data and the second real-time operating condition data respectively, the method comprising: collecting historical fault characteristics of the first construction equipment; collecting historical fault characteristics of the second building equipment; If the historical fault characteristics of the first construction equipment include the collaborative fault characteristics, and the historical fault characteristics of the second construction equipment include the collaborative fault characteristics; Collecting a first interval duration of the first construction equipment and collecting a second interval duration of the second construction equipment, wherein the interval duration is the interval duration from the last failure of the equipment; According to the fault prediction module, the probability of abnormal operating condition of the first building equipment is predicted based on the first interval duration and the collaborative fault characteristics, and the output is a first independent fault probability. The probability of abnormal operating condition of the second building equipment is predicted based on the second interval duration and the collaborative fault characteristics, and the output is a second independent fault probability. 3. The method according to claim 2, characterized in that a probability fusion calculation is performed according to the first independent failure probability and the second independent failure probability, and a fusion failure probability is output, and an expression of the probability fusion calculation includes: , in, Based on the first building equipment and second building equipment The probability of fusion failure, for the first construction equipment based on the first interval duration and the first independent failure probability of the collaborative failure signature D analysis, Where D is the collaborative fault characteristic A set of n is the number of features in the set, is the weight corresponding to the first independent fault probability obtained through integrated fusion model training; is a second independent failure probability of the second building equipment based on the second interval duration and the coordinated failure feature, is the weight corresponding to the second independent fault probability obtained through integrated fusion model training. 4. The method according to claim 3, characterized in that the method further comprises: If the historical fault characteristics of the first building equipment include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment do not include the collaborative fault characteristics; Predicting a first independent failure probability of the first construction equipment under the first interval time condition; A second initial failure probability corresponding to the second building equipment is obtained, and a probability fusion calculation is performed using the first independent failure probability and the second initial failure probability. The expression of the probability fusion calculation includes: , in, is a second initial failure probability corresponding to the second building equipment based on the collaborative failure feature, is the weight corresponding to the second initial fault probability obtained through integrated fusion model training. 5. The method according to claim 3, characterized in that the method further comprises: If the historical fault characteristics of the first building equipment do not include the collaborative fault characteristics, and the historical fault characteristics of the second building equipment include the collaborative fault characteristics; predicting a second independent failure probability of the second building equipment under the second interval time condition; A first initial failure probability corresponding to the first building equipment is obtained, and a probability fusion calculation is performed using the first initial failure probability and the second independent failure probability. The expression of the probability fusion calculation includes: , in, is a first initial failure probability corresponding to the first building equipment based on the collaborative failure feature, is the weight corresponding to the first initial fault probability obtained through integrated fusion model training. 6. The method according to claim 1, characterized in that, according to the first real-time working condition data and the second real-time working condition data, collaborative operation abnormality analysis is performed to obtain collaborative fault characteristics, and the method further comprises: Acquire historical collaborative working condition data of the first construction equipment and the second construction equipment, and perform collaborative operation abnormality analysis on the first real-time working condition data and the second real-time working condition data based on the historical collaborative working condition data to acquire collaborative fault characteristics; The collaborative fault feature is composed of an indicator whose data deviation is greater than or equal to a preset deviation. 7. The method according to claim 3, characterized in that the method further comprises: Obtain m initial candidate fusion models, and assign m different weight ratios to the m initial candidate fusion models ; Constructing a fusion sample set, the fusion sample comprising a failure probability training sample set based on the first construction equipment, a failure probability training sample set based on the second construction equipment, and a failure probability verification sample set of the first construction equipment and the second construction equipment working in collaboration; The m initial candidate fusion models are trained according to the fusion sample set, and an integrated fusion model is output. 8. The method according to claim 7, characterized in that the method further comprises: Training the m initial candidate fusion models according to the failure probability training sample set of the first building equipment and the failure probability training sample set of the second building equipment to obtain m failure probability output sample sets; Based on the comparison of the m fault probability output sample sets with the fault probability verification sample set, an optimized candidate fusion model is obtained, wherein the probability accuracy of the optimized candidate fusion model is greater than or equal to a preset accuracy; The weight ratio of the next round is adjusted according to the optimized candidate fusion model, and so on, to obtain m candidate fusion models after a preset number of iteration rounds; The m candidate fusion models are weighted and summed to obtain an integrated fusion model, and then the weight ratio of the integrated fusion model is output as and . 9. A fault prediction and analysis system for building equipment data fusion, characterized in that it is used to perform the steps of the method according to any one of claims 1 to 8, and the system comprises: A real-time working condition data acquisition unit, the real-time working condition data acquisition unit is used to collect first real-time working condition data of a first construction equipment and second real-time working condition data of a second construction equipment, wherein the first construction equipment and the second construction equipment are collaborative working equipment; A collaborative operation abnormality analysis unit, the collaborative operation abnormality analysis unit is used to perform collaborative operation abnormality analysis according to the first real-time working condition data and the second real-time working condition data to obtain collaborative fault characteristics; An independent fault probability prediction unit, the independent fault probability prediction unit is used to input the collaborative fault feature into a fault prediction module to perform fault prediction on the first real-time operating condition data and the second real-time operating condition data respectively, and obtain a first independent fault probability indicating that the first building equipment has failed, and a second independent fault probability indicating that the second building equipment has failed; A probability fusion calculation unit, wherein the probability fusion calculation unit is used to perform probability fusion calculation according to the first independent fault probability and the second independent fault probability, and output a fusion fault probability; A fault reminder unit is used to output fault reminder information when the fusion failure probability is greater than a preset fusion failure probability.